Recent advances and challenges in experiment-oriented polymer informatics

This review summarizes recent advances in experimental polymer chemistry supported by data science. The area of polymer informatics is rapidly growing based on cheminformatics, materials informatics, and data science platforms. Data-driven analyses, predictions, and suggestions for experimental polymer research are becoming more practical, and machine learning models can now predict various macromolecular properties with reasonable accuracy. At the same time, the limitations of current polymer informatics are being revealed. Developing appropriate treatments for higher-order structures and experimental procedures is critical to adequately process the hierarchical relationships of polymer systems. Recent attempts to treat this advanced information and future challenges in polymer informatics are discussed.